[
https://issues.apache.org/jira/browse/IGNITE-4488?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15773391#comment-15773391
]
Igor Sapego commented on IGNITE-4488:
-------------------------------------
Okay, query execution plan cached on the reducer-node. In case of C++ client
it's a client itself, in case of the ODBC node - it's node, on which
OdbcProcessor started and to which driver is connected.
> ODBC: First query execution takes much more time than any subsequent
> --------------------------------------------------------------------
>
> Key: IGNITE-4488
> URL: https://issues.apache.org/jira/browse/IGNITE-4488
> Project: Ignite
> Issue Type: Bug
> Components: odbc, platforms
> Affects Versions: 1.8
> Reporter: Igor Sapego
> Assignee: Igor Sapego
> Fix For: 2.0
>
>
> When we perform SQL query on small dataset just to measure latency, the very
> first call takes much more time to execute than the same subsequent requests
> (~70ms vs ~10ms).
> The same behavior can be observed with C++ client.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)